Anything that Can Go Wrong… Is to Be Equipped With a Failsafe System

 Anything that can go wrong, will go wrong, says Murphy's law. In order to avoid it, engineers design complex systems like jet controls and nuclear power plant failsafe systems.

Alexander Borgul, PhD student at the Department of Control Systems and Informatics develops automatic control algorithms to make such systems resistant to negative factors like turbulence which affects aircrafts or earthquakes that prevent static objects from proper operation.

According to Mr. Borgul, complex systems have the so-called saturation effect, which leads to large industrial accidents such as the Chernobyl disaster or the crash of F-22 Raptor.

“It means that a system behaves incorrectly even under negligible influence. When speaking of ascending and descending it is the crucial point because at those moments the system only “prepares” for operating mode. A misinterpreted signal can lead to the fall”, - says the researcher.

Most complex technical systems are tuned empirically. The engineers regulate the elements of a system manually, and then look at the results. There is still no magic button that can make everything work automatically and without troubles.

“My PhD project’s goal is to develop algorithms that will make the system resistant to external aggravators. For planes that would be turbulence, for static objects – earthquakes. It is essential that the necessary coefficients are to be taken automatically, allowing the system to function. Currently, I’m on the stage of testing my results using mathematic modelling”, - adds the scientist.

  Alexander Borgul shares that saturation effect can be erased if control activities are well-regulated. Adjusting algorithms are necessary for improving a system so that it continues to operate even if it obtains incorrect or incomplete information. Thanks to this, the system will manage to adjust to external conditions – like a living organism would.

“However, nothing comes without a price. Complex algorithms require more effective computational power, and using it lowers the system’s operation speed.”